Pilot Comparison of ⁶⁸Ga-RM2 PET and ⁶⁸Ga-PSMA-11 PET in Patients with Biochemically Recurrent Prostate Cancer

Patient Population

The local Radioactive Drug Research Committee, the local Institutional Review Board, and the Stanford Cancer Institute Scientific Review Committee approved the protocol. Written informed consent was obtained from all patients before they participated in the study. Inclusion criteria were age greater than 18 y at the time of radiotracer administration, a known diagnosis of prostate cancer, suspected recurrence based on biochemical data (PSA > 2 ng/mL), and ability to remain still during the imaging procedure (∼1 h). Exclusion criteria were metallic/conductive or electrically/magnetically active implants without MR-safe or MR-conditional labeling and standard contraindications to PET and MRI per the screening policy of our hospital (e.g., severe claustrophobia or radiation phobia).

Preparation of ⁶⁸Ga-PSMA-11

⁶⁸Ga-PSMA-11 was synthesized as previously reported (21). The precursor, Glu-NH-CO-NH-Lys(Ahx)-HBED-CC (DKFZ-PSMA-11, or PSMA-HBED), was obtained from ABX GmbH. All other reagents were of the highest grade, were purchased from a commercial supplier (Sigma, EMD Millipore), and were used as provided.

The radiosynthesis was performed on a fully automated device using sterile single-use cassettes (Modular Lab PharmTracer; Eckert & Ziegler Eurotope GmbH). ⁶⁸GaCl₃ was obtained from a ⁶⁸Ge/⁶⁸Ga generator (IGG-100 [1,850 MBq]; Eckert & Ziegler Isotope Products) by eluting the generator with 6 mL of 0.1 M HCl. The generator eluate was passed through a strong-cation-exchange cartridge to trap the ⁶⁸Ga ions for purification. Purified ⁶⁸GaCl₃ was eluted from the cartridge using a concentrated NaCl/HCl solution into a precharged reaction vial containing the precursor DKFZ-PSMA-11 (10 μg) in sodium acetate buffer (pH 4.5). The mixture was heated at 85°C for 180 s. Crude ⁶⁸Ga-PSMA-11 was diluted with saline and loaded onto a C18 light cartridge. The loaded cartridge was subsequently washed with saline and eluted with a 50% ethanol solution. The purified product was then diluted with saline and passed through a sterilizing membrane filter (0.2 μm) to afford the final formulated product (∼10 mL).

The identity and radiochemical purity of the product (⁶⁸Ga-PSMA-11) were evaluated by radio–high-performance liquid chromatography and radio–thin-layer chromatography. Estimated decay-corrected yield to the start of synthesis was 86.5% ± 4.1%, with a specific radioactivity of 70 ± 20 GBq/μmol (n = 10).

Preparation of ⁶⁸Ga-RM2

The precursor of ⁶⁸Ga-RM2, DOTA-4-amino-1-carboxymethylpiperidine-d-Phe-Gln-Trp-Ala-Val-Gly-His-Sta-Leu-NH₂ (DOTA-RM2), was obtained from ABX GmbH. A ⁶⁸Ga-labeling kit including eluent (concentrated NaCl/HCl solution), sodium acetate reaction buffer, ethanol (50% in water), and saline (0.9%, U.S. Pharmacopeia) was obtained from Eckert & Ziegler Eurotope GmbH. Phosphate buffer concentrate (1 M Na⁺, 0.6 M PO₄³⁻) for pH adjustment was obtained from B. Braun.

Purified ⁶⁸GaCl₃ was eluted from the strong-cation-exchange cartridge using a concentrated NaCl/HCl solution into a precharged reaction vial containing the precursor DOTA-RM2 (40 μg) in sodium acetate buffer (pH 4.5) and ethanol. The mixture was heated at 105°C for 400 s. Crude ⁶⁸Ga-RM2 was diluted with phosphate-buffered saline and loaded onto a C18 light cartridge. The loaded cartridge was subsequently washed with buffered saline and eluted with a 50% ethanol solution. The purified product was then diluted with phosphate-buffered saline and passed through a sterilizing membrane filter (0.2 μm) to afford the final formulated product (∼10 mL).

The identity and radiochemical purity of the product (⁶⁸Ga-RM2) were evaluated by radio–high-performance liquid chromatography and radio–thin-layer chromatography. Estimated decay-corrected yield to the start of synthesis was 86.1% ± 1.2%, with a specific radioactivity of 16 ± 1 GBq/μmol (n = 7).

PET/CT Protocol

No specific patient preparation such as fasting or hydration was requested for ⁶⁸Ga-PSMA-11 PET/CT scans. Whole-body PET/CT images were acquired from vertex to mid thighs with 8 bed positions and 3-min emission scans per bed position at 51–68 min (mean ± SD, 57.1 ± 5.9 min) after intravenous administration of the ⁶⁸Ga-PSMA-11. The administered activity was 130–144 MBq (mean ± SD, 137 ± 4 MBq). The images were reconstructed using an ordered-subset expectation maximization algorithm with 2 iterations and 32 subsets for the Discovery 600 scanner (GE Healthcare) (6 patients) and 2 iterations and 24 subsets for the Discovery 690 scanner (GE Healthcare) (1 patient). The CT acquisition was performed for attenuation correction, in helical mode, using 120 kV, 10 mAs, a 512 × 512 matrix, and an 867-mm field of view in 22.5 s. The CT amperage was reduced to 10 mAs, as requested by the local Institutional Review Board, to decrease the radiation exposure. The duration of the PET/CT examination ranged from 24 to 38 min (mean ± SD, 30.2 ± 4.9 min).

PET/MRI Protocol

No specific patient preparation such as fasting or hydration was required on the day of the ⁶⁸Ga-RM2 PET/MRI scans. Imaging (vertex to mid thighs) started at 42–51 min (mean ± SD, 48 ± 3 min) after injection of 133–152 MBq (mean ± SD, 137 ± 7 MBq) of ⁶⁸Ga-RM2 using a time-of-flight–enabled simultaneous PET/MRI scanner (Signa; GE Healthcare). The duration of the scan ranged from 39 to 85 min (mean ± SD, 66 ± 16 min) and, in addition to the difference in number of bed positions used related to patient height, also included the time to prepare the patients for the scans (e.g., positioning the body coils and defining the field of view). Lastly, patients who had not undergone prostatectomy received intravenous contrast material and underwent additional T1-weighted imaging over the prostate bed, contributing to the duration of the examination. The PET acquisition was performed in 3-dimensional mode at 4 min per bed position (89 slices per bed position) in 5–9 bed positions. An axial 2-point Dixon 3-dimensional T1-weighted spoiled gradient echo MR sequence (repetition time/first echo time/second echo time, 4.1/1.1/2.2 ms; field of view, 50 × 37.5 cm; matrix, 256 × 128; slice thickness/spacing, 5.2/2.6 mm; 120 images per slab; imaging time, 18 s) was acquired at each bed position and used to generate attenuation correction maps and to allow anatomic registration of the PET results.

PET images were reconstructed using ordered-subsets expectation maximization with 2 iterations and 28 subsets. Time-of-flight reconstructed images assumed a gaussian kernel of 400-ps width. The Dixon MRI sequence and the PET acquisition started at the same bed position and times, thus ensuring optimal temporal and regional correspondence between the MRI and PET data. For attenuation correction, the images were segmented into different tissue types with an anatomy-aware algorithm and were coregistered to a CT atlas in the head region (22).

Additional sequences acquired at each station included coronal T2-weighted single-shot fast spin echo imaging, diffusion-weighted imaging, and T1-weighted axial 2-point Dixon 3-dimensional spoiled gradient echo imaging, as previously described (22).

Image Analysis

Two board-certified nuclear medicine physicians with 10 and 8 y of experience in interpreting PET studies used MIMvista (version 6.2; MIMvista Corp.) to select organs throughout the body and used the region-of-interest (ROI) tool within the software to evaluate uptake. Circular ROIs, whose sizes (diameter, 10–30 mm) depended on the structure of interest, were drawn on transaxial ⁶⁸Ga-RM2 PET/MR images and transaxial ⁶⁸Ga-PSMA-11 PET/CT images with reference to anatomic structures confirmed by the MRI portion of the PET/MRI scan and the CT portion of the PET/CT scan, respectively. ROIs were analyzed for the frontal lobe cortex, cerebellar cortex, parotid gland, submandibular gland, thyroid, lung, ascending aorta as blood pool, liver, spleen, pancreas, small intestine, descending colon, kidney, bladder, gluteus maximus muscle, fat tissue of hip, right humerus, right femur, third cervical vertebra, ninth thoracic vertebra, third lumbar vertebra, and sacrum. To account for specific biodistribution, we added ROI measurements for esophagus and stomach for ⁶⁸Ga-RM2 and lacrimal grand and submandibular salivary gland for ⁶⁸Ga-PSMA-11. These ROIs were carefully positioned over the central portion of each structure depicted on the PET/MR or PET/CT image. For the aortic blood pool, a circular 10-mm-diamter ROI was placed centrally within the ascending aorta. A 30-mm-diameter ROI was placed on the right liver lobe.

For SUV measurements of focal ⁶⁸Ga-RM2 uptake outside the expected biodistribution, we reviewed the PET images with reference to the MRI data. Conversely, for SUV measurements of focal ⁶⁸Ga-PSMA-11 uptake outside the expected biodistribution, we reviewed the PET images with reference to the CT data. The PETedge tool of MIMvista was used to measure ⁶⁸Ga-RM2 and ⁶⁸Ga-PSMA-11 uptake outside the expected biodistribution.

Two board-certified radiologists (9 and 2 y of experience in interpreting body MRI studies) who did not know the results of PET or other studies evaluated the MR images for areas of abnormal signal or anatomic structures. Visual conspicuity against background on the diffusion-weighted images and presence of an anatomically corresponding abnormality on the T1- and T2-weighted images were the criteria for detecting a lesion on MRI.

Statistical Analysis

Data on time from injection to start of imaging, injected dosages of radiopharmaceuticals, SUV_mean, and SUV_max are presented as mean ± SD. The Mann–Whitney U test was used to evaluate focal uptake outside the expected biodistribution between ⁶⁸Ga-RM2 and ⁶⁸Ga-PSMA-11 (SUV_mean, SUV_max, and F/N ratio [ratio of SUV_mean and SUV_max to normal background, i.e., blood pool). The Pearson correlation coefficient was used to evaluate ⁶⁸Ga-RM2 and ⁶⁸Ga-PSMA-11 uptake. All statistical analyses were done with Stata 11 (Stata). Two-tailed P values of less than 0.05 were considered significant.